Virtual Data Augmentation, or VDA, is a framework for robustly fine-tuning pre-trained language model. Based on the original token embeddings, a multinomial mixture for augmenting virtual data is constructed, where a masked language model guarantees the semantic relevance and the Gaussian noise provides the augmentation diversity. Furthermore, a regularized training strategy is proposed to balance the two aspects.
Source: Virtual Data Augmentation: A Robust and General Framework for Fine-tuning Pre-trained ModelsPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Autonomous Driving | 1 | 16.67% |
Domain Generalization | 1 | 16.67% |
Semantic Segmentation | 1 | 16.67% |
Style Transfer | 1 | 16.67% |
Language Modelling | 1 | 16.67% |
Reinforcement Learning (RL) | 1 | 16.67% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |